60 likes | 137 Vues
Explore the issues in the patent analysis process and how rewording claims and descriptions is insufficient for patentability. Discover the drawbacks of word-based prior art searches and the costly nature of invalidating patents. Solutions include utilizing Amazon EC2 for data storage and opting for self-setup patent databases. Contact John Ellis for more information.
E N D
Analysis of Patents Rewording patent claims & descriptions doesn't make it patentable
Patent Analysis • USPTO Bulk Data downloadable via Google • Large storage arrays (4TB+) start at under $300 • Amazon EC2 reduces the barriers to entry • Code makes a difference - start programming
Problems with current process • Word-based prior art search is hit-and-miss, resulting in "slightly different" patents described in a way that sounds novel. • Invalidating patents is expensive. • Placing patents in primary classes that would be less likely to find prior art in the class.
Overused phrases in Patents "Heuristic Algorithm" "Over a network" "Searchable Database" "Plurality of" "converting the signal into one or more digital" "Distributed System Of" "communicate with the electronic device" "device is configured" "network based system"
Solutions • Designing systems with legacy compatibility with USPTO is not a major issue as the XML DTDs for patent data is well defined. • Solutions could partially be provided on Amazon EC2/S3 with common patent indexing and sharded archives was available for instances with minimal resources. • Virtually anyone with can setup their own patent database, to identify poor quality, abusive and potentially invalid patents.
Contact information John Ellis neurofog@gmail.com (510) 592-4153